Multilevel Latent Class Models with Dirichlet Mixing Distribution

Latent class analysis (LCA) and latent class regression (LCR) are widely used for modeling multivariate categorical outcomes in social sciences and biomedical studies. Standard analyses assume data of different respondents to be mutually independent, excluding application of the methods to familial and other designs in which participants are clustered. In this paper, we develop multilevel latent class model, in which subpopulation mixing probabilities are treated as random effects that vary among clusters according to a common Dirichlet distribution. We apply the Expectation-Maximization (EM) algorithm for model fitting by maximum likelihood (ML). This approach works well, but is computationally intensive when either the number of classes or the cluster size is large. We propose a maximum pairwise likelihood (MPL) approach via a modified EM algorithm for this case. We also show that a simple latent class analysis, combined with robust standard errors, provides another consistent, robust, but less efficient inferential procedure. Simulation studies suggest that the three methods work well in finite samples, and that the MPL estimates often enjoy comparable precision as the ML estimates. We apply our methods to the analysis of comorbid symptoms in the Obsessive Compulsive Disorder study. Our models' random effects structure has more straightforward interpretation than those of competing methods, thus should usefully augment tools available for latent class analysis of multilevel data.

Distribution and genetic variability among Campylobacter spp. isolates from different animal species and humans in Switzerland

In Switzerland, a national database with 1028 Campylobacter isolates from poultry, pigs, cats, dogs, cattle, humans, zoo animals and water has been created. The database contains the genetic fingerprint and background information of each Campylobacter isolate. Dominant species could be identified in the different sources with a majority of Campylobacter jejuni in poultry (73%), humans (79%), cattle (95%), zoo animals (40%) and water (100%), of Campylobacter coli in pigs (72%), and of Campylobacter upsaliensis/helveticus in cats and dogs (55%). The comparison of three genotyping methods, amplified fragment length polymorphism (AFLP), pulsed field gel electrophoresis and restriction fragment length polymorphism, revealed that AFLP allows discrimination between the different Campylobacter species and is the most appropriate method to distinguish specific strains within the same species. Genotyping analysis demonstrated that the Campylobacter population is heterogeneous among the different sources and that no dominant clone is spread in the country. Genotyping and the resulting database are useful tools to trace back future Campylobacter infections.

Phylogenetic relationship of equine Actinobacillus species and distribution of RTX toxin genes among clusters

Equine Actinobacillus species were analysed phylogenetically by 16S rRNA gene (rrs) sequencing focusing on the species Actinobacillus equuli, which has recently been subdivided into the non-haemolytic A. equuli subsp. equuli and the haemolytic A. equuli subsp. haemolyticus. In parallel we determined the profile for RTX toxin genes of the sample of strains by PCR testing for the presence of the A. equuli haemolysin gene aqx, and the toxin genes apxI, apxII, apxIII and apxIV, which are known in porcine pathogens such as Actinobacillus pleuropneumoniae and Actinobacillus suis. The rrs-based phylogenetic analysis revealed two distinct subclusters containing both A. equuli subsp. equuli and A. equuli subsp. haemolyticus distributed through both subclusters with no correlation to taxonomic classification. Within one of the rrs-based subclusters containing the A. equuli subsp. equuli type strain, clustered as well the porcine Actinobacillus suis strains. This latter is known to be also phenotypically closely related to A. equuli. The toxin gene analysis revealed that all A. equuli subsp. haemolyticus strains from both rrs subclusters specifically contained the aqx gene while the A. suis strains harboured the genes apxI and apxII. The aqx gene was found to be specific for A. equuli subsp. haemolyticus, since A. equuli subsp. equuli contained no aqx nor any of the other RTX genes tested. The specificity of aqx for the haemolytic equine A. equuli and ApxI and ApxII for the porcine A. suis indicates a role of these RTX toxins in host species predilection of the two closely related species of bacterial pathogens and allows PCR based diagnostic differentiation of the two.

Spatial Distribution of Cryptic Species Diversity in European Freshwater Amphipods (Gammarus fossarum) as Revealed by Pyrosequencing

In order to understand and protect ecosystems, local gene pools need to be evaluated with respect to their uniqueness. Cryptic species present a challenge in this context because their presence, if unrecognized, may lead to serious misjudgement of the distribution of evolutionarily distinct genetic entities. In this study, we describe the current geographical distribution of cryptic species of the ecologically important stream amphipod Gammarus fossarum (types A, B and C). We use a novel pyrosequencing assay for molecular species identification and survey 62 populations in Switzerland, plus several populations in Germany and eastern France. In addition, we compile data from previous publications (mainly Germany). A clear transition is observed from type A in the east (Danube and Po drainages) to types B and, more rarely, C in the west (Meuse, Rhone, and four smaller French river systems). Within the Rhine drainage, the cryptic species meet in a contact zone which spans the entire G. fossarum distribution range from north to south. This large-scale geographical sorting indicates that types A and B persisted in separate refugia during Pleistocene glaciations. Within the contact zone, the species rarely co-occur at the same site, suggesting that ecological processes may preclude long-term coexistence. The clear phylogeographical signal observed in this study implies that, in many parts of Europe, only one of the cryptic species is present.

Are cryptic host species also cryptic to parasites? Host specificity and geographical distribution of acanthocephalan parasites infecting freshwater Gammarus

Many parasites infect multiple host species. In coevolving host–parasite interactions, theory predicts that parasites should be adapted to locally common hosts, which could lead to regional shifts in host preferences. We studied the interaction between freshwater Gammarus (Crustacea, Amphipoda) and their acanthocephalan parasites using a large-scale field survey and experiments, combined with molecular identification of cryptic host and parasite species. Gammarus pulex is a common host for multiple species of Acanthocephala in Europe but, in Switzerland, is less common than two cryptic members of the Gammarus fossarum species complex (type A and type B). We found that natural populations of these cryptic species were frequently infected by Pomphorhynchus tereticollis and Polymorphus minutus. Four additional parasite species occurred only locally. Parasites were more common in G. fossarum type B than in type A. Infection experiments using several host and parasite sources confirmed consistently lower infection rates in G. pulex than in G. fossarum type A, suggesting a general difference in susceptibility between the two species. In conclusion, we could show that cryptic host species differ in their interactions with parasites, but that these differences were much less dramatic than differences between G. fossarum (type A) and G. pulex. Our data suggest that the acanthocephalans in Switzerland have adapted to the two most common Gammarus species in this region where host species frequencies differ from near-by regions in Europe.

Inheritance of female mating preference in a sympatric sibling species pair of Lake Victoria cichlids: implications for speciation

Female mate choice has often been proposed to play an important role in cases of rapid speciation, in particular in the explosively evolved haplochromine cichlid species flocks of the Great Lakes of East Africa. Little, if anything, is known in cichlid radiations about the heritability of female mating preferences. Entirely sympatric distribution, large ecological overlap and conspicuous differences in male nuptial coloration, and female preferences for these, make the sister species Pundamilia pundamilia and P. nyererei from Lake Victoria an ideally suited species pair to test assumptions on the genetics of mating preferences made in models of sympatric speciation. Female mate choice is necessary and sufficient to maintain reproductive isolation between these species, and it is perhaps not unlikely therefore, that female mate choice has been important during speciation. A prerequisite for this, which had remained untested in African cichlid fish, is that variation in female mating preferences is heritable. We investigated mating preferences of females of these sister species and their hybrids to test this assumption of most sympatric speciation models, and to further test the assumption of some models of sympatric speciation by sexual selection that female preference is a single-gene trait. We find that the differences in female mating preferences between the sister species are heritable, possibly with quite high heritabilities, and that few but probably more than one genetic loci contribute to this behavioural speciation trait with no apparent dominance. We discuss these results in the light of speciation models and the debate about the explosive radiation of cichlid fishes in Lake Victoria.

Matrix models for quantifying competitive intransitivity from species abundance data

In a network of competing species, a competitive intransitivity occurs when the ranking of competitive abilities does not follow a linear hierarchy (A > B > C but C > A). A variety of mathematical models suggests that intransitive networks can prevent or slow down competitive exclusion and maintain biodiversity by enhancing species coexistence. However, it has been difficult to assess empirically the relative importance of intransitive competition because a large number of pairwise species competition experiments are needed to construct a competition matrix that is used to parameterize existing models. Here we introduce a statistical framework for evaluating the contribution of intransitivity to community structure using species abundance matrices that are commonly generated from replicated sampling of species assemblages. We provide metrics and analytical methods for using abundance matrices to estimate species competition and patch transition matrices by using reverse-engineering and a colonization-competition model. These matrices provide complementary metrics to estimate the degree of intransitivity in the competition network of the sampled communities. Benchmark tests reveal that the proposed methods could successfully detect intransitive competition networks, even in the absence of direct measures of pairwise competitive strength. To illustrate the approach, we analyzed patterns of abundance and biomass of five species of necrophagous Diptera and eight species of their hymenopteran parasitoids that co-occur in beech forests in Germany. We found evidence for a strong competitive hierarchy within communities of flies and parasitoids. However, for parasitoids, there was a tendency towards increasing intransitivity in higher weight classes, which represented larger resource patches. These tests provide novel methods for empirically estimating the degree of intransitivity in competitive networks from observational datasets. They can be applied to experimental measures of pairwise species interactions, as well as to spatio-temporal samples of assemblages in homogenous environments or environmental gradients.

Tree size and fecundity influence ballistic seed dispersal of two dominant mast-fruiting species in a tropical rain forest

Seed production, seed dispersal, and seedling recruitment are integral to forest dynamics, especially in masting species. Often these are studied separately, yet scarcely ever for species with ballistic dispersal even though this mode of dispersal is common in legume trees of tropical African rain forests. Here, we studied two dominant main-canopy tree species, Microberlinia bisulcata and Tetraberlinia bifoliolata (Caesalpinioideae), in 25 ha of primary rain forest at Korup, Cameroon, during two successive masting events (2007/2010). In the vicinity of c. 100 and 130 trees of each species, 476/580 traps caught dispersed seeds and beneath their crowns c. 57,000 pod valves per species were inspected to estimate tree-level fecundity. Seed production of trees increased non-linearly and asymptotically with increasing stem diameters. It was unequal within the two species’ populations, and differed strongly between years to foster both spatial and temporal patchiness in seed rain. The M. bisulcata trees could begin seeding at 42–44 cm diameter: at a much larger size than could T. bifoliolata (25 cm). Nevertheless, per capita life-time reproductive capacity was c. five times greater in M. bisulcata than T. bifoliolata owing to former’s larger adult stature, lower mortality rate (despite a shorter life-time) and smaller seed mass. The two species displayed strong differences in their dispersal capabilities. Inverse modelling (IM) revealed that dispersal of M. bisulcata was best described by a lognormal kernel. Most seeds landed at 10–15 m from stems, with 1% of them going beyond 80 m (<100 m). The direct estimates of fecundity significantly improved the models fitted. The lognormal also described well the seedling recruitment distribution of this species in 121 ground plots. By contrast, the lower intensity of masting and more limited dispersal of the heavier-seeded T. bifoliolata prevented reliable IM. For this species, seed density as function of distance to traps suggested a maximum dispersal distance of 40–50 m, and a correspondingly more aggregated seedling recruitment pattern ensued than for M. bisulcata. From this integrated field study, we conclude that the reproductive traits of M. bisulcata give it a considerable advantage over T. bifoliolata by better dispersing more seeds per capita to reach more suitable establishment sites, and combined with other key traits they explain its local dominance in the forest. Understanding the linkages between size at onset of maturity, individual fecundity, and dispersal capability can better inform the life-history strategies, and hence management, of co-occurring tree species in tropical forests.

The spatial distribution of water in the inner coma of Comet 9P/Tempel 1: Comparison between models and observations

The near nucleus coma of Comet 9P/Tempel 1 has been simulated with the 3D Direct Simulation Monte Carlo (DSMC) code PDSC++ (Su, C.-C. [2013]. Parallel Direct Simulation Monte Carlo (DSMC) Methods for Modeling Rarefied Gas Dynamics. PhD Thesis, National Chiao Tung University, Taiwan) and the derived column densities have been compared to observations of the water vapour distribution found by using infrared imaging spectrometer on the Deep Impact spacecraft (Feaga, L.M., A’Hearn, M.F., Sunshine, J.M., Groussin, O., Farnham, T.L. [2007]. Icarus 191(2), 134–145. http://dx.doi.org/10.1016/j.icarus.2007.04.038). Modelled total production rates are also compared to various observations made at the time of the Deep Impact encounter. Three different models were tested. For all models, the shape model constructed from the Deep Impact observations by Thomas et al. (Thomas, P.C., Veverka, J., Belton, M.J.S., Hidy, A., A’Hearn, M.F., Farnham, T.L., et al. [2007]. Icarus, 187(1), 4–15. http://dx.doi.org/10.1016/j.icarus.2006.12.013) was used. Outgassing depending only on the cosine of the solar insolation angle on each shape model facet is shown to provide an unsatisfactory model. Models constructed on the basis of active areas suggested by Kossacki and Szutowicz (Kossacki, K., Szutowicz, S. [2008]. Icarus, 195(2), 705–724. http://dx.doi.org/10.1016/j.icarus.2007.12.014) are shown to be superior. The Kossacki and Szutowicz model, however, also shows deficits which we have sought to improve upon. For the best model we investigate the properties of the outflow.

Antibiotic Susceptibility and Sequence Type Distribution of Ureaplasma Species Isolated from Genital Samples in Switzerland.

Antibiotic resistance in Ureaplasma urealyticum/Ureaplasma parvum and Mycoplasma hominis is an issue of increasing importance. However, data regarding the susceptibility and, more importantly, the clonality of these organisms are limited. We analyzed 140 genital samples obtained in Bern, Switzerland, in 2014. Identification and antimicrobial susceptibility tests were performed by using the Mycoplasma IST 2 kit and sequencing of 16S rRNA genes. MICs for ciprofloxacin and azithromycin were obtained in broth microdilution assays. Clonality was analyzed with PCR-based subtyping and multilocus sequence typing (MLST), whereas quinolone resistance and macrolide resistance were studied by sequencing gyrA, gyrB, parC, and parE genes, as well as 23S rRNA genes and genes encoding L4/L22 ribosomal proteins. A total of 103 samples were confirmed as positive for U. urealyticum/U. parvum, whereas 21 were positive for both U. urealyticum/U. parvum and M. hominis. According to the IST 2 kit, the rates of nonsusceptibility were highest for ciprofloxacin (19.4%) and ofloxacin (9.7%), whereas low rates were observed for clarithromycin (4.9%), erythromycin (1.9%), and azithromycin (1%). However, inconsistent results between microdilution and IST 2 kit assays were recorded. Various sequence types (STs) observed previously in China (ST1, ST2, ST4, ST9, ST22, and ST47), as well as eight novel lineages, were detected. Only some quinolone-resistant isolates had amino acid substitutions in ParC (Ser83Leu in U. parvum of serovar 6) and ParE (Val417Thr in U. parvum of serovar 1 and the novel Thr417Val substitution in U. urealyticum). Isolates with mutations in 23S rRNA or substitutions in L4/L22 were not detected. This is the first study analyzing the susceptibility of U. urealyticum/U. parvum isolates in Switzerland and the clonality outside China. Resistance rates were low compared to those in other countries. We hypothesize that some hyperepidemic STs spread worldwide via sexual intercourse. Large combined microbiological and clinical studies should address this important issue.

Environmental correlates of species rank − abundance distributions in global drylands

Theoretical models predict lognormal species abundance distributions (SADs) in stable and productive environments, with log-series SADs in less stable, dispersal driven communities. We studied patterns of relative species abundances of perennial vascular plants in global dryland communities to: (i) assess the influence of climatic and soil characteristics on the observed SADs, (ii) infer how environmental variability influences relative abundances, and (iii) evaluate how colonisation dynamics and environmental filters shape abundance distributions. We fitted lognormal and log-series SADs to 91 sites containing at least 15 species of perennial vascular plants. The dependence of species relative abundances on soil and climate variables was assessed using general linear models. Irrespective of habitat type and latitude, the majority of the SADs (70.3%) were best described by a lognormal distribution. Lognormal SADs were associated with low annual precipitation, higher aridity, high soil carbon content, and higher variability of climate variables and soil nitrate. Our results do not corroborate models predicting the prevalence of log-series SADs in dryland communities. As lognormal SADs were particularly associated with sites with drier conditions and a higher environmental variability, we reject models linking lognormality to environmental stability and high productivity conditions. Instead our results point to the prevalence of lognormal SADs in heterogeneous environments, allowing for more evenly distributed plant communities, or in stressful ecosystems, which are generally shaped by strong habitat filters and limited colonisation. This suggests that drylands may be resilient to environmental changes because the many species with intermediate relative abundances could take over ecosystem functioning if the environment becomes suboptimal for dominant species.